Evidence for a role of the hepatic endocytic compartment in the modulation of the extracellular matrix

The presence of fibronectin in rat liver endocytic compartment was investigated using biochemical and immunological approaches. Three endosome subfractions were separated from postmitochondrial lysosome supernatants using sucrose and isoosmotic Nycodenz gradients. Using these endosomes the presence of fibronectin in the "early" and "late" endosome subfractions and in a receptor-enriched fraction was demonstrated by Western blot analysis. Furthermore, immunofluorescence studies using an anti-rat fibronectin antiserum and an anti-rat endosome antiserum showed similar patterns of staining in frozen liver sections. The results indicate that the components of the extracellular matrix extend into the endocytic compartment and suggest that fibronectin is internalised in a manner similar to that of some plasma membrane proteins.     

Functional polarity of the rat hepatocyte surface membrane. Isolation and characterization of plasma-membrane subfractions from the blood-sinusoidal, bile-Canalicular and contiguous surfaces of the hepatocyte.

1. Six rat liver plasma-membrane subfractions of different density and morphological, enzymic and chemical properties were prepared from homogenates by a combination of differential, rate-zonal and density-gradient centrifugation. They consisted of three vesicular 'light' subfractions of density 1.12-1.13 and three 'heavy' subfractions of density 1.16-1.18 containing membrane strips and intercellular junctions. 2. All six subfractions contained a basal adenylate cyclase activity. One of the 'light' subfractions that showed the highest glucagon-stimulated adenylate cyclase activity was identified as deriving form the blood-sinusoidal face of the hepatocyte. This subfraction, unlike the others, was contaminated by Golgi components, as indicated by its morphological properties and the presence of galactosyl- and sialyl-transferase activities. 3. All the six subfractions showed high activities of the following plasma-membrane marker enzymes: 5'-nucleotidase, alkaline phosphodiesterase (nucleotide pyrophosphatase), alkaline phosphatase, leucine naphthylamidase and Mg2+-activated adenosine triphosphatase. A 'light' subfraction that showed the highest specific activities of all the above marker enzymes, but lacked a glucagon-stimulated adenylate cyclase activity, was identified as deriving from the bile-canalicular face of the hepatocyte. 4. The 'heavy' subfractions, which showed generally the lowest activities of the above plasma-membrane enzyme markers, and were characterized by the presence of desmosomes and gap junctions, were taken to originate from the contiguous faces of the hepatocyte. 5. The protein composition of the six subfractions was generally similar, as shown by polyacrylamide-gel electrophoresis. Differences in the amounts of various protein and glycoprotein bands among the subfractions correlated with their morphology, enzymic composition and sialic acid content. 6. Hormonal and histochemical evidence supporting the identification of a bile-canalicular subfraction, a blood-sinusoidal subfraction and contiguous-face subfractions is discussed.     

Hepatic endosome fractions contain an ATP-driven proton pump.

Endosome fractions were isolated from rat liver homogenates on the basis of the subcellular distribution of circulating ligands, e.g. 125I-asialotransferrin internalized by hepatocytes by a receptor-mediated process. The distribution of endocytosed 125I-asialotransferrin 1-2 min and 15 min after uptake by liver and a monensin-activated Mg2+-dependent ATPase activity coincided on linear gradients of sucrose and Nycodenz. The monensin-activated Mg2+-ATPase was enriched relative to the liver homogenates up to 60-fold in specific activity in the endosome fractions. Contamination of the endosome fractions by lysosomes, endoplasmic reticulum, mitochondria, plasma membranes and Golgi-apparatus components was low. By use of 9-aminoacridine, a probe for pH gradients, the endosome vesicles were shown to acidify on addition of ATP. Acidification was reversed by addition of monensin. The results indicate that endosome fractions contain an ATP-driven proton pump. The ionophore-activated Mg2+-ATPase in combination with the presence of undegraded ligands in the endosome fractions emerge as linked markers for this new subcellular organelle.     

T-system membranes in microsomal fractions of crayfish muscles: identification and differences in protein composition

Specific binding of 3H-ouabain and ruthenium red (RR) to membranes of T-tubules in crayfish muscles was used to identify the subfraction containing vesicles originating from the T-system. The microsomal fraction was prepared by differential centrifugation, and subfractions were separated in continuous or discontinuous sucrose density gradients. 3H-ouabain binding was estimated by scintillation counting; RR binding was examined by electron microscopy. The light subfraction was identified using both methods as that containing vesicles of T-tubules. Protein separation by SDS-electrophoresis revealed marked differences between the subfraction containing vesicles of T-tubules and other subfractions, the most distinctive feature being the presence of a protein of Mr 46,000 predominantly in the light subfraction.     

Phospholipid, cholesterol, polypeptide and glycoprotein composition of hepatic endosome subfractions.

The composition of hepatic endosome subfractions was compared directly with that of plasma membranes and Golgi-apparatus fractions. The neutral lipid and phospholipid composition of two endosome subfractions separated on Nycodenz gradients from a parent endosome fraction was similar to that of plasma membranes. The phospholipid/cholesterol ratios and the sphingomyelin contents were high, as in plasma membranes. However, the phosphatidylserine content was low. Endosomal subfractions contained a simpler polypeptide profile than did plasma membranes. However, a large number of glycoproteins were common to both fractions. Two endosome-specific glycoproteins, of Mr 59 000 and 38 000, were identified. Sialic acid was present at concentrations higher than in plasma membranes. The results indicate that endosomal membranes have a similar composition to plasma membranes that probably reflects their functional interaction during endocytosis and receptor recycling.     

Relations between plasma membrane and lysosomal membrane. 2. Quantitative evaluation of plasma membrane marker enzymes in the lysosomes

We have quantified, in cultured rat fibroblasts, the association to the lysosomal membrane of two classical plasma membrane markers, 5'-nucleotidase and alkaline phosphodiesterase I. To isolate highly purified lysosomal preparations, lysosomes were loaded with horseradish peroxidase (2-h cell uptake, 16-h chase) and isolated by isopycnic centrifugation in linear Percoll gradients, followed by a 3,3'-diaminobenzidine-induced density shift in sucrose gradients. Purified lysosomal preparations contained up to 50% of N-acetyl-beta-glucosaminidase of the homogenate. This lysosomal enzyme was enriched 33-fold in the most purified preparations. In the electron microscope, these preparations appeared to be highly purified and only contained organelles filled with diaminobenzidine reaction products. Analysis of purified preparations indicates that 0.5-0.8% of 5'-nucleotidase, but as much as 10.9-14.3% of alkaline phosphodiesterase I activities of the homogenate, are associated with lysosomes. After freezing-thawing, these activities remained essentially membrane-associated. The larger value obtained for alkaline phosphodiesterase I could not be ascribed to other lysosomal enzymes, as no such activity was detected at acidic pH. These two plasma membrane markers are thus unevenly distributed in the lysosomal compartment.     

Fractionation of endocytic vesicles and glucose-transporter-containing vesicles in rat adipocytes.

We subfractionated intracellular vesicles from rat adipocytes in order to examine the subcellular distribution of endocytic vesicles or endosomes with respect to insulin-regulatable glucose-transporter (GT)-containing vesicles [James, Lederman & Pilch (1987) J. Biol. Chem. 262, 11817-11824]. Vesicles mediating fluid-phase endocytosis sedimented as a single major peak of greater density than the single distinct peak of GT-containing vesicles. This difference was also apparent during cellular insulin exposure and after insulin removal. Endocytosis of insulin and IGF (insulin-like growth factor) II was also examined. In sucrose gradients, IGF II-containing vesicles were less dense than those containing internalized insulin. Receptor-mediated endocytic vesicles were distinct from fluid-phase endocytic vesicles, but overlapped with the GT-containing vesicles. Vesicles containing internalized ligand were further fractionated by agarose-gel electrophoresis after various times of internalization. At least three different vesicle subpopulations containing the iodinated ligands were resolved after 5 min of internalization. Endocytic vesicles containing rapidly internalized insulin (1.5 min at 37 degrees C) consistently co-migrated with GT-containing vesicles. These data indicate that fluid-phase and receptor-mediated endocytosis occur via different pathways in adipocytes. Furthermore, whereas the intracellular GT-containing vesicles are distinct from fluid-phase vesicles, a rapidly labelled pool of insulin-containing vesicles consistently co-fractionated with GT-containing vesicles when separation techniques based on size, density and charge were used. This suggests that the insulin receptor may directly interact with the intracellular GT-containing vesicles after insulin-induced endocytosis.     

Electronegative LDLs from familial hypercholesterolemic patients are physicochemically heterogeneous but uniformly proapoptotic

A highly electronegative fraction of human plasma LDLs, designated L5, has distinctive biological activity that includes induction of apoptosis in bovine aortic endothelial cells (BAECs). This study was performed to identify a relationship between LDL density, electronegativity, and biological activity, namely, the induction of apoptosis in BAECs. Plasma LDLs from normolipidemic subjects and homozygotic familial hypercholesterolemia subjects were separated into five subfractions, with increasing electronegativity from L1 to L5, and into seven subfractions according to increasing density, D1 to D7. L1 to L5 were also separated according to density, and D1 to D7 were separated according to charge. The density profiles of L1 to L5 were similar (maximum density = 1.030 +/- 0.002 g/ml). Induction of apoptosis by all seven density subfractions was confined to the highly electronegative fraction, L5, and within each density subfraction the magnitude of apoptosis correlated with the L5 content. Electronegative LDL is heterogeneous with respect to density and composition, and induction of apoptosis is more strongly associated with LDL electronegativity than with LDL size or density.     

Fractionation of liver plasma membranes prepared by zonal centrifugation

1. Plasma membranes were isolated from crude nuclear sediments from mouse and rat liver by a rate-dependent centrifugation through a sucrose density gradient contained in the ;A' type zonal rotor. 2. The membranes were further purified by isopycnic centrifugation, and characterized enzymically, chemically and morphologically. 3. When the plasma-membrane fraction of sucrose density 1.17g/cm(3) was dispersed in a tight-fitting homogenizer, two subfractions of densities 1.12 and 1.18 were obtained by isopycnic centrifugation. 4. The light subfraction contained 5'-nucleotidase, nucleoside diphosphatase, leucine naphthylamidase and Mg(2+)-stimulated adenosine triphosphatase activities at higher specific activities than unfractionated membranes. The heavy subfraction was deficient in the above enzymes but contained higher Na(+)+K(+)-stimulated adenosine triphosphatase activity. 5. The light subfraction contained twice as much phospholipid and cholesterol, and three times as much N-acetylneuraminic acid relative to unit protein weight as the heavy subfraction. Polyacrylamide-gel electrophoresis indicated differences in protein composition. 6. Electron microscopy showed the light subfraction to be vesicular. The heavy subfraction contained membrane strips with junctional complexes in addition to vesicles.     

Enzymatic and chemical analyses of pig platelet membrane subfractions isolated by zonal centrifugation.

1. A mixed membrane fraction prepared from pig platelets was subfractionated, using the "B 14" zonal rotor, into two distinct subpopulations of membrane vesicles, each associated with a different phosphodiesterase activity. 2. The lighter subfraction (MI) was enriched 7-8 fold with bis-(p-nitrophenyl) phosphate phosphodiesterase activity and the denser subfraction (MII) showed a similar degree of enrichment of 5'dTMP-p-nitrophenyl ester phosphodiesterase activity. 3. Assays for other enzyme activities revealed slight enrichement (approx. 2 fold) of acid phosphatase, 3'-dTMP-p-nitrophenyl ester phosphodiesterase and beta-glucuronidase activities in MI, and beta-galactosidase in MII. Cyclic AMP phosphodiesterase, lactate dehydrogenase and N-acetyl-beta-glucosaminidase showed negligible activity in both MI and MII, and succinate dehydrogenase activity could not be detected in either subfraction. 4. Chemical analyses of the membrane subfractions demonstrated that MI contained approx. twice as much cholesterol, phospholipid, sialic acid and hexosamine per unit weight of protein than MII. These results are consistent with our previously reported observations from surface-labelling experiments, which indicated that MI was derived principally from the platelet surface-exposed membranes and that MII was probably intracellular in origin. 5. Analysis of the membrane polypeptides by sodium dodecyl sulphate-polyacrylamide gel electrophoresis revealed the presence of 12-15 components, in each subfraction, in the mol. wt. range 12000-200000, including a prominent band of approx. mol. wt. 46000, which has beeen identified to be actin. Qualitative as well as possible quantitative differences were apparent in that MII contained three components in addition to those present in MI. 6. Analysis of the periodate-Schiff staining components by sodium dodecyl sulphate-polyacrylamide gel electrophoresis demonstrated the presence of 4 major glycoproteins in both subfractions with apparent mol. wt. ranging from approx. 95000 to 150000; in addition two minor components were also present. Further, a very fast-migrating band, which did not stain with Coomassie blue, was observed in both MI and MII and probably represents lipid material.     

Preparation of plasma-membrane subfractions from isolated rat hepatocytes.

1. Rat livers were dissociated into their constituent cells by perfusion through the portal vein with a medium containing collagenase, and hepatocytes separated from non-parenchymal cells. 2. It is shown that the procedure described by Wisher & Evans [(1975) Biochem. J. 146, 375-388] for preparation of plasma membranes from liver tissue when applied to isolated hepatocytes also yielded subfractions of similar morphology and marker-enzyme distribution. 3. Thus the distribution of alkaline phosphodiesterase, 5'-nucleotidase and the basal and glucagon-stimulated adenylate cyclase among two 'light' vesicular and one 'heavy' junction-containing plasma-membrane subfractions paralleled that reported for tissue-derived plasma-membrane subfractions. 4. Increased recoveries and specific activities of plasma-membrane marker enzymes were obtained when soya-bean trypsin inhibitor was included in the collagenase-containing perfusion media used to dissociate the liver. 5. Polyacrylamide-gel-electrophoretic analysis of the corresponding plasma-membrane subfractions prepared from liver tissue and isolated hepatocytes were generally similar. 6. The results indicate that the functional polarity of the hepatocyte's plasma membrane is retained after tissue dissociation. The damage occurring to plasma-membrane ectoenzymes by the collagenase-perfusion procedure is discussed.     

CYTOCHEMICAL LOCALIZATION OF 5''-NUCLEOTIDASE IN SUBCELLULAR FRACTIONS ISOLATED FROM RAT LIVER : I. The Origin of 5''-Nucleotidase Activity in Microsomes

A procedure has been developed for the cytochemical localization of 5'-nucleotidase in isolated, unfixed, rat liver microsomes. Membranes were incubated with adenosine 5'-phosphate (5'-AMP) and Pb(NO₃)₂ and then isolated on sucrose density gradients: all the phosphate released was recovered with the membranes by this procedure. Adenosine 2'-phosphate (2'-AMP) and adenosine 3', 5'-cyclic phosphate (3',5'-AMP) were shown to be competitive inhibitors, but not substrates, for purified 5'-nucleotidase and were employed to determine the specificity of the cytochemical reaction. It was found that the incubation conditions for the cytochemical assay did not affect the specificity of 5'-nucleotidase. Microsomes incubated as controls with Pb²⁺, or Pb²⁺ and 2'-AMP or 3',5'-AMP were of the same density, although slightly denser than microsomes incubated without Pb²⁺, and were unassociated with lead precipitate when examined by electron microscopy; microsomes incubated with Pb²⁺ and 5'-AMP were much denser and were stained heterogeneously with lead phosphate when examined by electron microscopy. Precipitates formed artificially from Pb²⁺ and inorganic phosphate did not resemble the reaction product. Microsomes were, therefore, separated on sucrose gradients and the subfractions were examined cytochemically. Lead precipitates were associated with the majority of rough-surfaced vesicles, and the reaction product was distributed heterogeneously in all fractions. Vesicles which stained like the membranes of the bile canaliculi in isolated plasma membranes were observed in the lightest subfraction. The reaction product was localized on the outside surface of the microsomal membranes, and was solubilized by low concentrations of ethylenediaminetetraacetic acid. It is concluded that 5'-nucleotidase is present in the endoplasmic reticulum and that the microsome fraction contains, in addition, vesicles derived from the plasma membrane.     

SUBCELLULAR DISTRIBUTION OF NORADRENALINE IN GUINEA-PIG CEREBRAL CORTEX AND SPLEEN

Abstract— The distribution of noradrenaline (NA) in subcellular fractions of guinea-pig cerebral cortex and spleen was determined by differential and density gradient centrifugation. Of the primary fractions, the microsomal fraction from both tissues was enriched in NA, that of the spleen having the higher specific activity. Microsomal fractions were therefore placed on gradients and NA determined in the subfractions since these fractions appeared suitable preparations in which to search for discrete populations of vesicles. So that the non-occluded micro-particulate bound noradrenaline (MPBNA) content of gradient subfractions could be measured, [³H]NA was used to control for the diffusion and or adsorption of free NA, and occluded lactate dehydrogenase was used to estimate the amount of entrapped MPBNA and soluble NA. Non-occluded MPBNA on gradients from microsomal fractions of cerebral cortex formed a single peak mainly in subfraction F (0.6-0.8 m -sucrose). Spleen microsomal fractions, however yielded two peaks of MPBNA. one in sub-fractions D to G (0.4-1.0 m -sucrose) and the other in sub-fraction J (1.4 m -sucrosc); electron microscopy showed that the latter subfraction contained large vesicles.
Since there were unexpectedly small amounts of MPBNA in microsomal subfractions D and E of cerebral cortex, the synaptosome fraction was investigated. Following water treatment of synaptosomes. MPBNA formed a peak in subfraction E (0.4-0.6 m -sucrose) with smaller amounts in subfractions D and F (0.4 and 0.6 0.8 m -sucrose).     

Characterization of rat liver endosomal fractions. In vivo activation of insulin-stimulable receptor kinase in these structures

A protocol employing discontinuous sucrose gradient centrifugation was developed to prepare light mitochondrial (L) and Golgi fraction endosomes from simultaneously prepared parent L and microsomal fractions. As judged by the concentration of labeled hormone postinjection, L intermediate and heavy endosome subfractions were 40- to 175-fold purified and Golgi intermediate and heavy endosome subfractions were 30- to 45-fold purified. On electron microscopy, L endosomal fractions contained a predominance of lipoprotein-filled vesicles and were less heterogeneous than corresponding Golgi endosomal fractions. All endosomal fractions were enriched in receptors for insulin and prolactin but binding sites for the former were more broadly distributed in other subfractions than those for the latter. On Percoll gradient centrifugation, L endosomal fractions yielded one peak (rho 1.057) corresponding to the heavier of two peaks seen in Golgi endosomal fractions. The protein composition of high density L and Golgi endosomes, as assessed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, was similar. The bulk of marker enzymes assayed did not migrate with the endosomal components. Combined acid phosphatase cytochemistry and electron microscope radioautography established that about 80% of the L endosomes contained no acid phosphatase. By affinity labeling and immunological titration with insulin receptor antibody, insulin receptors were identical in L and Golgi endosomes. Insulin-stimulable receptor kinase was demonstrable in both L and Golgi endosome fractions. Following in vivo insulin administration, the insulin receptor kinase in both L and Golgi endosomes was significantly activated. This activated state was not inhibited by a large excess of antiserum to insulin and thus not due to insulin contaminating the partially purified receptor preparation. These observations are compatible with the maintenance and/or initiation of hormone-dependent phosphorylations intracellularly.     

Structural and functional polarity of canalicular and basolateral plasma membrane vesicles isolated in high yield from rat liver

A method has been developed for routine high yield separation of canalicular (cLPM) from basolateral (blLPM) liver plasma membrane vesicles of rat liver. Using a combination of rate zonal floatation (TZ- 28 zonal rotor, Sorvall) and high speed centrifugation through discontinuous sucrose gradients, 9-16 mg of cLPM and 15-28 mg of blLPM protein can be isolated in 1 d. cLPM are free of the basolateral markers Na+/K+-ATPase and glucagon-stimulatable adenylate cyclase activities, but are highly enriched with respect to homogenate in the "canalicular marker" enzyme activities leucylnaphthylamidase (48-fold), gamma-glutamyl-transpeptidase (60-fold), 5'-nucleotidase (64-fold), alkaline phosphatase (71-fold), Mg++-ATPase (83-fold), and alkaline phosphodiesterase I (116-fold). In contrast, blLPM are 34-fold enriched in Na+/K+-ATPase activity, exhibit considerable glucagon-stimulatable adenylate cyclase activity, and demonstrate a 4- to 15-fold increase over homogenate in the various "canalicular markers." cLPM have a twofold higher content of sialic acids, cholesterol; and sphingomyelin compared with blLPM. At least three canalicular-(130,000, 100,000, and 58,000 mol wt) and several basolateral-specific protein bands have been detected after SDS PAGE of the two LPM subfractions. Specifically, the immunoglobin A-binding secretory component is restricted to blLPM as demonstrated by immunochemical techniques. These data indicate virtually complete separation of basolateral from canalicular LPM and demonstrate multiple functional and compositional polarity between the two surface domains of hepatocytes.     

Identification of an endosome-specific antigen.

We have used successive density gradient centrifugation with vesicles prepared from a human hepatoma Hep G2 post nuclear supernatant to obtain a highly enriched preparation of early endosomes. A monoclonal antibody (8E4) raised against this early endosome preparation recognizes a single polypeptide highly enriched in light vesicle membranes. The antigen has a molecular weight of 195 kDa by SDS-PAGE in the presence or absence of a reducing agent. Western blot analysis shows that the 8E4 antigen is detectable only in light vesicle membranes and not among heavy membranes, whole cytosol, or nuclear pellet proteins. The 8E4 antigen appears to be an integral membrane protein as it is precipitated by Triton X-114. The distribution of the 8E4 antigen in a Nycodenz density gradient fractionation of light vesicle membranes is identical to the distribution of 125I-ASOR-labeled early endosomes but distinct from the distribution of the plasma membrane enzyme, alkaline phosphodiesterase. In addition, incubation of cells with a horseradish peroxidase-transferrin conjugate followed by 3,3'-diaminobenzidine cytochemistry specifically quenches 8E4 antigen detection by protein dot blot analysis. These data strongly suggest that the 8E4 antigen is an integral membrane protein primarily located in endocytic vesicles.     

Relation between insulin resistance and fast-migrating LDL subfraction as characterized by capillary isotachophoresis

The proportion of the electronegative low density lipoprotein [LDL(-)] subfraction, which is atherogenic, is increased in type 2 diabetes but is not reduced by glycemic control. Therefore, we evaluated the ability of a new technique, capillary isotachophoresis (cITP), to quantify charge-based LDL subfractions and examined the relation between insulin resistance and the cITP fast-migrating (f) LDL levels. Seventy-five 10-year-old boys were included. The two cITP LDL subfractions, fLDL and major LDL subfractions, were proportional to the LDL protein content within the range of 0.1-0.8 mg/ml LDL protein. Levels of cITP fLDL were positively correlated with triglyceride (TG) levels and negatively correlated with LDL size. Insulin resistance as assessed by the homeostasis model assessment (HOMA-IR) was positively correlated (P < 0.01) with cITP fLDL levels (r = 0.41). The relation between HOMA-IR and cITP fLDL levels depended on TG levels but was independent of body mass index and LDL size. cITP lipoprotein analysis is an accurate and sensitive method for quantifying charge-based LDL subfractions in human plasma, and insulin resistance is related to cITP fLDL independent of LDL size.     

Endosome pH measured in single cells by dual fluorescence flow cytometry: rapid acidification of insulin to pH 6

The acidification of various ligands was measured on a cell by cell basis for cell suspensions by correlated dual fluorescence flow cytometry. Mouse 3T3 cells were incubated with a mixture of fluorescein- and rhodamine-conjugated ligands, and the ratio of fluorescein and rhodamine fluorescence was used as a measure of endosome pH. The calibration of this ratio by both fluorometry and flow cytometry is described. Dual parameter histograms of average endosome pH per cell versus amount of internalization were calculated from this data, for samples in the absence and presence of chloroquine added to neutralize acidic cellular vesicles. The kinetics of acidification of insulin were measured and compared with previous results obtained with the chloroquine ratio technique. Rapid acidification of internalized ligand was observed both for insulin, which was mostly internalized via nonspecific pathways, and for alpha 2-macroglobulin, which was mainly internalized by specific receptor-mediated endocytosis. The average pH observed for internalized insulin was less than pH 6 within 10 min after addition of insulin. At 30 min, the average pH began to decrease to approximately pH 5, presumably because of fusion of endosomes with lysosomes.     

Isolation and partial characterization of the luminal plasmalemma of microvascular endothelium from rat lungs.

This paper describes a procedure for isolating in high yield and at a high degree of purity the endothelial luminal plasmalemma from the microvasculature of the rat lung. The procedure relies on the modification of the density of the luminal plasmalemma obtained by coating it by perfusion in situ first, with cationized colloidal silica and then with Na polyacrylate. These steps generate a strongly adhering coat to the luminal plasmalemma that resists tissue homogenization to yield, upon repeated centrifugation through Nycodenz density gradients, a nearly homogeneous fraction of coated luminal plasmalemmal fragments still carrying their associated plasmalemmal vesicles. The fraction is enriched in the luminal plasmalemmal antigen, angiotensin converting enzyme, contains gp60, an antigen expected to occur on both plasmalemmal domains, is not enriched in either alkaline phosphatase or 5'-nucleotidase activity and is free of the mitochondrial and endoplasmic reticulum antigens so far tested. This procedure, that can be extended--in principle--to any vascular bed, obviates the use of cultured cells for studying the biochemistry of the endothelium, at least as far as the luminal endothelial plasmalemma is concerned.     

Glycerylphosphorylcholine phosphodiesterase in rat liver. Subcellular distribution and localization in plasma membranes

1. A simple new assay for glycerylphosphorylcholine phosphodiesterase is described, in which radioactive glycerylphosphorylcholine is used as substrate and the reaction products are separated by adsorption on an anion-exchange resin. 2. Rat liver subcellular fractions contained both particulate (58%) and soluble (42%) glycerylphosphorylcholine phosphodiesterase. Both activities released free choline from glycerylphosphorylcholine. 3. The particulate glycerylphosphorylcholine phosphodiesterase was recovered mainly in the nuclear and microsomal fractions and showed a distribution similar to those of 5'-nucleotidase and alkaline phosphodiesterase I, both of which are constituents of the liver plasma membrane. 4. During purification of plasma membranes glycerylphosphorylcholine phosphodiesterase, 5'-nucleotidase and alkaline phosphodiesterase I showed largely similar behaviour, indicating that glycerylphosphorylcholine phosphodiesterase is also localized in liver plasma membranes. Slight differences in the distributions of these three enzymes in density-gradient separations are discussed in relation to the possibility that they are unevenly distributed on different areas of the cell surface. 5. The differences between glycerylphosphorylcholine phosphodiesterase and alkaline phosphodiesterase I indicate that these two activities are not functions of a single enzyme. 6. The glycerylphosphorylcholine phosphodiesterase of liver plasma membranes has a pH optimum of 8.5 and a K(m) for glycerylphosphorylcholine of 0.95mm. It is inhibited by EDTA and fully reactivated by a variety of bivalent cations (and Fe(3+)).